ChatGPT was trained on a vast corpus of text data from the internet, including websites, books, articles, and more. Its knowledge is up-to-date as of October 2023. It does not have awareness of events or developments that occurred after this date. The training process involves unsupervised learning to predict the next word in a sentence and fine-tuning with human feedback to improve response quality and adherence to guidelines.
ChatGPT’s Training Data and Knowledge Cutoff
1 minuteread
